elsevier.com/locate/yexnr

Experimental Neurology 1

Regular Article

FAS deficiency reduces apoptosis, spares axons and improves function

after spinal cord injury

S. Casha, W.R. Yu, M.G. Fehlings *

Spinal Program, Krembil Neuroscience Center, Toronto Western Hospital, 399 Bathurst St., Toronto, Ontario, Canada M5T 2S8

Department of Surgery, Institute of Medical Science, University of Toronto, 399 Bathurst St., Toronto, Ontario, Canada M5T 2S8

Received 1 April 2005; revised 15 July 2005; accepted 25 August 2005

Available online 3 October 2005

Abstract

After spinal cord injury (SCI), apoptosis of neurons and oligodendrocytes is associated with axonal degeneration and loss of neurological

function. Recent data have suggested a potential role for FAS death receptor-mediated apoptosis in the pathophysiology of SCI. In this study, we

examined the effect of FAS deficiency on SCI in vitro and in vivo. FASLpr/lpr mutant mice and wildtype background-matched mice were subjected

to a T5–6 clip compression SCI, and complementary studies were done in an organotypic slice culture model of SCI. Post-traumatic apoptosis in

the spinal cord, which was seen in neurons and oligodendrocytes, was decreased in the FAS-deficient mice both in vivo and in vitro particularly in

oligodendrocytes. FAS deficiency was also associated with improved locomotor recovery, axonal sparing and preservation of oligodendrocytes

and myelin. However, FAS deficiency did not result in a significant increase in surviving neurons in the spinal cord at 6 weeks after injury, likely

reflecting the importance of other cell death mechanisms for neurons. We conclude that inhibition of the FAS pathway may be a clinically

attractive neuroprotective strategy directed towards oligodendroglial and axonal preservation in the treatment of SCI and neurotrauma.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Spinal cord injury; FAS; Apoptosis; Oligodendrocyte; Neuroprotection

Introduction

A number of studies have linked apoptosis to spinal cord

injury (SCI) (Casha and Fehlings, 2001; Crowe et al., 1997;

Emery et al., 1998;Katoh et al., 1996; Li et al., 1996, 1999; Liu

et al., 1997; Lou et al., 1998; Yong et al., 1998). In these reports,

early apoptosis of neural cells including neurons is followed by

a delayed wave of predominantly oligodendroglial programmed

cell death in degenerating white matter tracts (Crowe et al.,

1997; Emery et al., 1998; Li et al., 1999; Liu et al., 1997;

Shuman et al., 1997). Studies of apoptosis in white matter after

dorsal cordotomy or after transection suggest that glial

apoptosis occurs, at least in part, as a consequence of axonal

degeneration (Abe et al., 1999; Warden et al., 2001). The loss of

trophic support derived by the oligodendrocyte from the axon

likely results in activation of programmed cell death. However,

0014-4886/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.expneurol.2005.08.020

* Corresponding author. Spinal Program, Krembil Neuroscience Center,

Toronto Western Hospital, 399 Bathurst St., Toronto, Ontario, Canada M5T

2S8. Fax: +1 416 603 5298.

E-mail address: Michael.fehlings@uhn.on.ca (M.G. Fehlings).

the presence of activated microglia in contact with apoptotic

oligodendrocytes after SCI suggests that this interaction may

also activate cell death programs in the oligodendrocyte

(Shuman et al., 1997). Secondary axonal degeneration may

then follow as is seen in models of multiple sclerosis (Bjartmar

et al., 1999) and in myelin-associated glycoprotein deficiency

(Yin et al., 1998). It is thus logical to postulate that decreasing

apoptosis in both oligodendrocytes and neurons may improve

neurological outcome after SCI.

FAS (CD95 or APO1) is known to interact with cytoplasmic

FAS-Associated Death Domain protein (FADD) and, upon FAS

ligand (FasL) binding, allows activation of caspase 8 leading to

apoptosis (Siegel et al., 2000). This receptor exhibits changes

in expression which may implicate it in the pathophysiology of

SCI. Li and colleagues showed that white matter expression of

FAS and FasL is lost within 1 day at the site of injury while

increasing in segments adjacent to the injury site (Li et al.,

2000a). Zurita et al. found an increase in the number of cells

expressing FAS increases within 72 h of SCI (Zurita et al.,

2001). We previously demonstrated that FAS expression,

caspase 8 and 3 activation and apoptosis occur in a temporally

96 (2005) 390 – 400

www.

S. Casha et al. / Experimental Neurology 196 (2005) 390–400 391

similar fashion after SCI (Casha and Fehlings, 2001). Demjen

et al. demonstrated that FasL neutralizing antibodies improved

neurological outcome after transaction SCI (Demjen et al.,

2004). In addition, FAS signaling has been implicated in death

of motor neurons after spinal ischemia and axotomy of the

facial nerve (Matsushita et al., 2000; Sakurai et al., 1998;

Ugolini et al., 2003). Given these observations, we hypothesize

that deficiency of FAS receptor will result in decreased

oligodendroglial cell death and improved behavioral and

histological outcome after SCI.

In this article, we compared in vivo and in vitro SCI in

C57BL/6-FASLpr/lpr mice and wildtype C57BL/6 background-

matched mice. These lpr mice are deficient in FAS expression

due to a retro-transposon insertion in the second intron of the

FAS gene leading to a premature stop codon and a non-

functional gene product (Nakatsuru et al., 1999). Phenotypical-

ly, these mice appear normal at birth but develop in a strain-

dependent manner a lymphoproliferative disorder at ages

greater than 18 weeks (Vidal et al., 1998). We used animals

with a C57Bl/6J background at 8 weeks of age to avoid this

potentially confounding feature, which develops later and less

severely in this strain (Vidal et al., 1998). In addition, an in vitro

organotypic culture model was also used to confirm findings in

the absence of systemically mediated inflammation.

Materials and methods

In vivo spinal cord injury

SCI was performed in 8-week-old female B6.MRL-Fas-lpr

mice (lpr) and C57BL/6 background-matched mice (Jackson

Laboratory, Bar Harbor, Maine) using the FEJOTAi clip

compression model at T5/6 (Joshi and Fehlings, 2002a). Under

(1.5–2%) halothane and N2O (1 L/min) anesthetic, a modified

aneurysm clip (7.8 g closing force) was applied rapidly for 1

min to the spinal cord after laminectomy. After injury, animals

were housed at 25–27-C. All protocols were in accordance

with the Canadian Council of Animal Care policies and were

approved by the animal care committee of the University

Health Network.

Organotypic slice culture model

We employed an in vitro organotypic slice culture model of

SCI (Krassioukov et al., 2002). Briefly, 350 Am slices from

T2–T10 spinal cord of 4-week-old mice were generated

(McIlwain tissue chopper, Harvard apparatus, Saint Laurent,

Quebec) and incubated in culture medium (50% MEM with

Earl’s salts and glutamine, 25% Hanks balanced salt solution,

25% horse serum, 20 mM HEPES, 6 mg/ml d-glucose, pH 7.2)

at 36-C in 5% CO2. At 14 days, the slices were injured using a

guided weight dropped onto the center of the slice (0.137 g

piston, 1.2 mm diameter, 5 mm height) (Fig. 3) 30 min after

addition of propidium iodide (10 AM) to the media. Uptake of

propidium iodide was imaged using a confocal microscope

(Bio-Rad MRC 600), and pixels with grayscale intensity >127

at 24, 48 and 72 h were counted (ImagePro plus software,

Media Cybernetics, Silver Springs, Maryland). Lpr and wild-

type counts were compared by 2-way ANOVA (n = 50 slices).

Western blotting

Five millimeters of spinal cord centered at the injury site or

the organotypic spinal cord cultures were homogenized in 10

volumes of 4-C 50 mM Tris 4 mM EDTA, pH 7.4 with 1 AMpepstatin, 100 AM leupeptin, 100 AM phenylmethylsulfonyl-

fluoride and 10 Ag/ml aprotinin. Twenty micrograms of protein

was subjected to SDS-PAGE followed by electroblotting onto

nitrocellulose. Blots were blocked (1 h) with 1% nonfat dry

milk 0.1% Tween 20 Tris-buffered saline and incubated for 1

h with primary antiserum (NF200 (Sigma, Oakville, Ontario),

MAP2 (Chemicon, Temecula, California), CNPase (Sternber-

ger Monoclonals, Lutherville, Massachusetts), caspase 8 or

caspase 3 (Santa Cruz Biotechnology, Santa Cruz, California)

and for 1 h with horseradish-peroxidase-conjugated secondary

antiserum. Immunoreactive bands were visualized by chemi-

luminescence (ECL, Amersham, Arlington Heights, Illinois),

and band density was compared (Gelpro plus software, Media

Cybernetics, Silver Springs, Maryland). Actin immunoblotting

and Coomassie Blue staining were compared to ensure equal

loading of samples (Schumacher et al., 1999). All Western

blots were performed three times using independent samples.

In situ terminal-deoxy-transferase mediated dUTP nick end

labeling (TUNEL)

Injured lpr and wildtype mice (n = 4) were perfusion-fixed

with 4% paraformaldehyde 0.1 M phosphate buffer at days 1

and 3 post-injury. Serial 10 Am paraffin spinal cord sections at

500 Am intervals were stained with hematoxylin, eosin and

Luxol Fast Blue to identify the injury epicenter. The TUNEL

assay was applied using the Apotag kit (Intergen, New York,

New York) to sections at the injury epicenter and at 1000 Am.

TUNEL-positive nuclei were counted through the entire cross-

section and compared by Student’s t test.

Wildtype mouse spinal cord sections 500 Am from the injury

epicenter 7 days after injury and sections generated from in

vitro injured spinal cord tissue were double labeled with cell-

specific antisera and TUNEL. Oligodendrocyte stained sections

were blocked for 1 h with 5% milk 0.5% BSA and incubated

with anti-CNPase antiserum (1:500) or anti-CC1 antiserum

(1:20, Calbiochem, San Diego, California) at 4-C overnight.

Neuron-stained sections were blocked for 1 h in 20% goat

serum and incubated with anti-NeuN antiserum (1:50, Chemi-

con, Temecula, California) overnight at 4-C. FITC-conjugatedsecondary antiserum was applied for 1 h followed by TUNEL

as above.

Immunohistochemistry

Wildtype mouse spinal cord sections 500 Am from the injury

epicenter 7 days after injury were double labeled with cell-

specific antisera and FAS. Sections were blocked with goat

serum and incubated overnight with anti-FAS antiserum (Santa

S. Casha et al. / Experimental Neurology 196 (2005) 390–400392

Cruz Biotechnology, Santa Cruz, CA) (1:50 in 20 mM PBS).

Sections were exposed to horseradish-peroxidase-conjugated or

fluorescent (Texas Red or FITC)-conjugated secondary antise-

rum. Aminoethyl carbazole (AEC) was used to develop the

peroxidase reaction. Oligodendrocyte-stained sections were

blocked for 1 h with 5% milk 0.5% BSA and incubated with

anti-CNPase antiserum (1:500) or anti-CC1 antiserum (1:20,

Calbiochem, San Diego, California) at 4-C overnight. Neuron-

stained sections were blocked for 1 h in 20% goat serum and

incubated with anti-NeuN antiserum (1:50, Chemicon, Teme-

cula, California) overnight at 4-C. Astrocytes were stained withanti-GFAP serum (Sigma, Oakville, Ontario) (1:200 in 20 mM

PBS) overnight at 4-C after blocking with goat serum. FITC-

conjugated or Texas-Red-conjugated secondary antiserum was

then applied for 1 h.

Sections 2000 Am from the injury epicenter at 7 days post-

injury and sham-injured controls were immunostained with anti-

CNPase antiserum or NeuN antiserum as above (n = 3/group).

A random 40� field was selected, and the number of

immunostained cells was counted. Lpr and wildtype counts

were compared by Student’s t test.

NeuN-positive neurons were also counted in paraffin cross-

sections at 500, 1000, 1500 and 2000 Am from the injury

epicenter and at 6 weeks post-injury. Counts in lpr and

wildtype animals were compared by 2-way ANOVA (variables

FAS, genome and distance n = 6/group).

In vivo behavioral scoring

Injured wildtype (n = 6) and lpr (n = 6) mice were scored

using the modified BBB locomotor behavioral score for mice

by two blinded individuals at weekly intervals following SCI

Fig. 1. Post-traumatic apoptosis in the mouse spinal cord. Immunohistochemical exa

cell-specific markers confirmed apoptosis in oligodendrocytes (A–C) and neurons (

fields double labeled with NeuN (neuronal marker) and CNPase (oligodendrocytes m

merged (scale bar, 20 Am).

(Basso et al., 1995; Joshi and Fehlings, 2002a). Scores in

wildtype and mutant mice were compared by 2-way ANOVA.

Fluorogold retrograde tracing

Six weeks after injury, the animals underwent application of

a 4% Fluorogold-soaked gelfoam pledget to the rostral surface

of a T10 transection. Five days later, the animals were

perfusion fixed with 4% paraformaldehyde, and serial 40 Amparaffin sections of the brain stem were prepared. The number

of Fluorogold containing neurons in every second section

through the vestibular, raphe, red and reticular nuclei was

counted in a blinded manner and compared by Student’s t test.

The Fluorogold application site was examined to ensure that

the area of diffusion did not approach the injury site.

Retrograde tracing in non-injured control animals and T5/6

transection control animals was also quantified.

Luxol Fast Blue staining

Serial 10 Am paraffin cross-sections of spinal cord 6 week

post-injury were stained with 0.1% Luxol Fast Blue (Sigma,

Oakville, Ontario) in 95% ethanol, 0.5% acetic acid and

destained using 0.05% LiCO3. Stained sections at 250 Amintervals around the injury epicenter were digitally imaged, and

the area of Luxol Fast Blue staining was determined as the

percent of pixels with grayscale value >127 on the inverse

image (ImagePro Plus software, Media Cybernetics, Silver

Springs, MD). Mean lpr and wildtype areas were compared by

2-way ANOVA. Pairwise comparisons were made using t test

with Bonferroni correction. Uninjured T5/6 spinal cord

sections were also compared (n = 3 mutant and n = 4 wildtype).

mination of frozen sections 500 Am from the injury epicenter with TUNEL- and

D–F) 7 days after in vivo SCI. TUNEL is shown in panels A and D. The same

arker) are shown in panels B and E respectively. In panels C and F, the fields are

S. Casha et al. / Experimental Neurology 196 (2005) 390–400 393

Statistical analyses

The specific statistical comparisons used in each experiment

are indicated in the above methods. P < 0.05 was used to reject

the null hypothesis and to indicate statistical significance. All

tests were 2-tailed.

Results

Apoptotic cell death following murine SCI

As a prelude to investigations in FAS-deficient mice, we

first confirmed the occurrence of apoptosis in this species using

the TUNEL method (Figs. 1A, D, 4A, B). Double labeling

immunohistochemistry with TUNEL- and cell-specific markers

confirmed the occurrence of apoptosis in both oligodendro-

cytes and neurons (Fig. 1). Furthermore, double labeling

immunohistochemistry with FAS and cell markers demonstrat-

ed that FAS is expressed in oligodendrocytes, neurons and

some astrocytes in the mouse (Fig. 2). TUNEL and FAS double

Fig. 2. Cellular distribution of FAS expression in wildtype mice. FAS immunohisto

labeling at 7 days post-injury focusing on a region 500 Am from the injury epicen

(arrow, D–F). FAS expression was rarely seen in astrocytes (arrow, G–I). FAS was p

in the left and center column as indicated, while the merged image is presented on

labeling confirmed the occurrence of FAS-positive apoptotic

cells after SCI (Fig. 2).

The effect of FAS deficiency on cell death after in vitro SCI

In our in vitro slice SCI model (Figs. 3A, B), we found that

the extent of cell death, as assayed by propidium iodide uptake,

was significantly lower in lpr preparations than wildtype

preparations at 48 and 72 h (2-way ANOVA: P < 0.0001 for

FAS genome and time; pairwise comparisons: P < 0.05 at 48

and 72 h) (Figs. 3C–E). Given that propidium iodide does not

distinguish apoptotic from necrotic death, we examined

caspase 8 and 3 activation in this model by Western blotting

for the p20 band of the activated caspases. Caspase 8 and 3 are

the apical and terminal caspases respectively in the FAS-

mediated apoptotic pathway. We found greater activation of

both caspases in wildtype slices than in lpr slices at 72 h post-

injury (Fig. 3F), suggesting that less programmed cell death

occurs in FAS-deficient mice after in vitro SCI and in the

absence of a systemic inflammatory response.

chemistry was combined with cell marker immunohistochemistry and TUNEL

ter. We found FAS expression in oligodendrocytes (arrow, A–C) and neurons

resent on TUNEL-positive cells (arrow, J–L). Individual labels are represented

the right (scale bar, 20 Am).

Fig. 3. Decreased cell death in injured FAS-deficient spinal cord slice preparations. (A) Organotypic slice culture preparation stained for myelin with Luxol Fast

Blue. Slices were injured by a weight drop technique ((B) the injury device positioned using a micomanipulator (1), a 0.137 g piston (3) guided by a sleeve (2)

dropped from 5 mm height on to the center of the slice). The mean area of propidium iodide uptake (pixel counts) in FAS lpr mutant slices was significantly lower (*)

than in wildtype counterparts at 48 and 72 h after injury (E) (representative propidium iodide fluorograms at 72 h are shown for (C) FAS-deficient and (D) wildtype

preparations). Densitometry of Western blots revealed significantly decreased activated caspase 8 and 3 p20 subunit (*) (F).

S. Casha et al. / Experimental Neurology 196 (2005) 390–400394

Acute cell death after in vivo SCI in FAS-deficient and wildtype

mice

In order to examine the role of FAS in a more clinically

relevant model, we next compared apoptosis in lpr and

wildtype animals after in vivo clip compression SCI. Compar-

ison of TUNEL counts (Figs. 4A–C) showed no difference

between lpr and wildtype mice at 24 h, however, there was a

significant decrease in the number of apoptotic cells in lpr mice

at 72 h (t test: P = 0.034). Similarly, by Western blotting,

caspase 3 activation was greater in wildtype animals than in lpr

animals at day 7 post-injury (t test: P = 0.002) (Fig. 5D) but

was not significantly different at days 1 and 2 post-injury or in

uninjured controls. Thus, FAS deficiency resulted in reduced

post-traumatic apoptosis in the spinal cord.

NF200 is a neuronal cytoskeletal protein that has been shown

to degrade locally in the spinal cord following injury. Its loss

reflects loss of neuronal soma or axons (Schumacher et al.,

2000). Degradation of NF200 seen in wildtype animals within

24 h of the injury was largely attenuated in the lpr mice (Fig. 5A).

MAP2 is a neuronal protein largely restricted to the neuronal

soma. On Western blotting, MAP2 signal was decreased in

wildtype mice when compared to mutant at 7 days (t test: P =

0.70 at day 1, P = 0.049 at day 7) (Fig. 5B). These data suggest

that FAS deficiency influences preservation of neuronal

elements early after injury in the spinal cord. However, counts

of NeuN-positive neurons done at 7 days (t test: P = 0.442)

(Fig. 5C) post-injury did not reveal any difference between lpr

and wildtype (similarly no difference in neuron counts was seen

at 6 weeks; ANOVA P = 0.989, data not shown). This suggests

that while FAS deficiency influences neuronal structural

proteins this does not impact neuronal survival at the injury site.

Previous work by our group and others emphasized the

importance of oligodendrocyte apoptosis after SCI (Casha and

Fehlings, 2001). We therefore examined CNPase preservation

after SCI. We observed increased CNPase signal on Western

blotting (Fig. 5D) and increased numbers of CNPase-positive

cells (t test: P = 0.027) (Fig. 5E) 7 days following SCI in lpr

Fig. 4. Reduced post-traumatic apoptosis in the spinal cord of FAS-deficient mice after in vivo SCI. Counts of TUNEL-labeled nuclei revealed no difference between

FAS mutant and wildtype animals at 24 h but were significantly decreased (*) in FAS-deficient animals at 72 h (C). TUNEL-positive nuclei were not uniformly

distributed throughout the cross-section of the spinal cord but were more numerous in the vicinity of the injury center ((A, B) TUNEL in dorsolateral white matter of

wildtype and mutant respectively at 72 h, scale bar: 50 Am). Western blotting for activated caspase 3 did not reveal a significant difference in uninjured animals or at

days 1 and 2 post-injury, however, at day 7, a significant increase (*) in caspase 3 activation was seen in wildtype animals (D). A representative Western blot for

caspase 3 in wildtype and mutant spinal cord at day 7 is illustrated in (E).

S. Casha et al. / Experimental Neurology 196 (2005) 390–400 395

mice. Oligodendrocyte counts in sham-injured wildtype mice

were not significantly different than those seen in lpr mice (t test:

P = 0.196).

In vivo behavioral and neuroanatomical comparisons of SCI in

FAS-deficient and wildtype mice

We compared spinal-cord-injured lpr mice to wildtype

background-matched mice (n = 6/group) for 6 weeks after

injury using the modified BBB locomotor behavioral score for

mice at weekly intervals (Basso et al., 1995; Joshi and Fehlings,

2002a). There were no animal mortalities in either group, and no

morbidities other than the neurological deficit caused by SCI

were seen. The behavioral recovery observed over time in the lpr

mice was significantly better than that seen in wildtype animals

(2-way ANOVA: P = 0.001 for FAS genome, P < 0.001 for time

post-injury) (Fig. 6). The mean BBB scores reached a plateau at

8 in the wildtype group (representing hindlimb lateral sweeping

or plantar placement of the hindlimbs with no weight support)

and 10 in the lpr animals (representing weight-supported plantar

steps seen in up to 50% of steps but lacking coordination with

forelimb movements).

Six weeks after injury, we re-anesthetized the animals and

administered the retrograde tracer Fluorogold caudal to the

injury. After 5 days of retrograde uptake, we counted labeled

neurons of the vestibular, raphe, red and reticular nuclei in

cross-sections through the brainstem. These bulbospinal

projections contribute significantly to recovery of hindlimb

locomotor function in mice after SCI (Joshi and Fehlings,

2002b). We found a significant increase in the number of

brainstem neurons retrogradely labeled by Fluorogold, indicat-

ing a greater preservation of descending axons across the injury

site in lpr animals (t tests: red nucleus P = 0.0006, reticular

nuclei P = 0.001, raphe nuclei P = 0.044, vestibular nuclei P =

0.014) (Fig. 7). Retrograde tracing in non-injured control

animals revealed no background difference between wildtype

and lpr mice. T5/6 transection control animals did not exhibit

any retrogradely labeled neurons in the brainstem.

At the time of perfusion fixation for Fluorogold, we removed

a 6 mm spinal cord segment centered at the injury and stained

serial paraffin cross-sections with the myelin-selective pigment

Luxol Fast Blue. We compared the area of staining within the

spinal cord images. This experiment revealed a significant

improvement in white matter myelin preservation at the injury

site in lpr animals (2-way ANOVA: P = 0.02 for FAS genome,

P < 0.001 for distance from injury epicenter) (Fig. 8). Uninjured

T5/6 spinal cord, wildtype and mutant sections showed no

difference in Luxol Fast Blue staining.

Fig. 5. Preservation of oligodendrocyte and neuronal markers after in vivo SCI in FAS-deficient mice. Western blotting showed an attenuation of NF200 (arrow)

degradation in FAS-deficient animals at 24 h (lpr 1) when compared to wildtype animals (+/+ 1). lpr 0 and +/+ 0 represent respective uninjured controls (A). Western

blotting for MAP2 in the spinal cord revealed gradual loss in wildtype animals through 7 days (+/+ days 1–7) that was attenuated in mutant animals (lpr days 1–7)

(B). Counts of NeuN-labeled neurons 7 days after injury were unchanged in FAS mutant mice compared to wildtype (C). CNPase Western blotting at 7 days after

SCI revealed significantly increased (*) CNPase preservation in FAS mutant mice (D). Counts of CNPase labeled oligodendrocytes 7 days after injury were

significantly increased (*) in FAS mutant mice compared to wildtype (E).

S. Casha et al. / Experimental Neurology 196 (2005) 390–400396

These findings indicate that FAS deficiency is associated

with greater myelin preservation, improved axonal survival and

significantly improved locomotor function after traumatic SCI.

Discussion

In this study, we found that FAS deficiency results in

decreased apoptosis and selective preservation of oligoden-

drocytes but not neurons after SCI. Deficiency of the FAS

receptor results in improved neurological recovery after SCI,

with improved axonal survival and myelin preservation. These

findings implicate the FAS death receptor pathway in apoptotic

secondary mechanisms after SCI particularly in oligodendro-

cytes. They provide evidence that strategies to decrease

delayed glial cell death after SCI may be associated with

axonal preservation and are potentially efficacious neuropro-

tective strategies.

FAS deficiency results in axonal and oligodendroglial

preservation after SCI

We observed apoptosis in both neurons and oligodendrocytes

after SCI in the mouse. Decreased apoptosis in FAS-deficient lpr

animals was evidenced by TUNEL and caspase activation.

Decreased NF200 and MAP2 degradation provides evidence of

improved structural integrity of the axons and neurons in the

vicinity of the injury, although no effect on neuronal survival

was seen in vivo (Schumacher et al., 1999, 2000). Furthermore,

greater axonal preservation in chronically injured lpr animals

was evidenced by retrograde Fluorogold uptake. In addition, we

found evidence of enhanced oligodendrocyte and myelin

preservation in lpr mice. The effect of FAS deficiency in

preserving oligodendrocytes is consistent with the observation

of several laboratories that post-traumatic apoptosis in the spinal

cord occurs largely in oligodendrocytes (Casha and Fehlings,

2001; Crowe et al., 1997; Emery et al., 1998; Katoh et al., 1996;

Li et al., 1996). In addition, Demjen et al. (2004) used FasL

inhibiting antibodies to show improved neurological outcome

and evidence of improved sprouting of corticospinal tract axons.

These authors also demonstrated evidence of oligodendrocyte

preservation with FasL neutralization.

Some strains of FAS-deficient lpr mice exhibit a lympho-

proliferative disorder. To minimize this effect, we used C57BL

background strain mice less than 8 weeks old (maximum

survival 14 weeks). These mice exhibit this disorder minimally

and late in life (beyond 18 weeks) (Vidal et al., 1998). Thus, it

is unlikely that differences in the immune system played a

significant role in our results. Moreover, our in vitro model

Fig. 7. Improved axonal preservation after SCI in FAS-deficient mice. Six

weeks after injury, a Fluorogold impregnated gelfoam pledget was applied to

the rostral surface of a T10 transection site (A1). Five days later, retrograde-

labeled neurons in the brainstem (A3) (thus having intact axons across the

injury site (A2)) were counted. Mean count T standard error of the mean is

presented (D). Differences were statistically significant at each nucleus counted

(*). Representative Fluorogold fluorograms of the red nucleus from FAS-

deficient and wildtype animals respectively are shown in panels B and C (scale

bar, 100 Am). Inset in panel A is a longitudinal section through the Fluorogold

introduction site at day 5 (scale 1 mm) demonstrating that the area of passive

diffusion of the Fluorogold extended well short of the injury site (approxi-

mately 5 mm rostral).

Fig. 6. Improved behavioral outcome after SCI in FAS-deficient mice.

Behavioral recovery following SCI determined using the BBB locomotor

score was significantly better in the lpr animals compared to wildtype (graph of

mean score T standard error of the mean vs. time post-injury). Selected

functional definitions of the BBB score relevant to the graph are presented.

S. Casha et al. / Experimental Neurology 196 (2005) 390–400 397

excluded the systemic inflammatory response suggesting that

the differences observed in mutant animals were indeed due to

the effect of FAS in the spinal cord. However, we recognize

that our in vitro preparation includes cell types other than

neurons and glia and that with time differences in cell death

and proliferation may alter the cellular composition of the slice

from that seen in vivo. Thus, while useful in confirming in vivo

findings, this model is not intended to provide strong

independent mechanistic evidence.

FAS deficiency results in improved behavioral and

neuroanatomical outcome after SCI

In our chronic in vivo study comparing lpr animals to

wildtype background-matched mice, we found behavioral

improvement, improvement in the number of intact axons

across the injury site and a decrease in white matter loss.

The beneficial outcome seen suggests that the FAS death

receptor cascade is a potentially useful target in the

treatment of SCI and that amelioration of delayed cell death

may afford significant neurological improvement. In addi-

tion, as most apoptotic cell loss occurs within 2 weeks of

injury, the therapeutic window for intervention in this

condition is longer than previous studies have explored

(Casha and Fehlings, 2001).

Yoshino et al. also compared SCI in FAS-deficient lpr

mice and wildtype controls (Yoshino et al., 2004). Similar to

our study, they demonstrated fewer TUNEL-positive cells,

improved behavioral outcome and a decrease in the

histological evidence of injury in mutant mice. However,

in contrast to our study, they observed a peak in the number

of TUNEL-positive cells at day 1 and a resolution of the

difference between mutant and wildtype by day 4. Further-

more, they identified these cells to be largely neurons. Our

data indicated little difference in neuron counts and a

predominant effect of FAS mutation in oligodendrocyte

preservation. In addition, we observed increased caspase

Fig. 8. Improved myelin preservation after SCI in FAS-deficient mice. Panels A and B are representative micrographs at the injury epicenter from a FAS-deficient

and wildtype animal respectively stained with Luxol Fast Blue (scale 50 Am). The mean area of myelin staining at the injury site was significantly greater (*) in FAS-

deficient animals -o- than in their wildtype counterparts -.- (C).

S. Casha et al. / Experimental Neurology 196 (2005) 390–400398

activation through day 7. The reason for the differences

between these two studies is unclear however may be due to

difference in the background mouse strain used or differ-

ences in the injury model. While early neuronal apoptosis

has been observed previously (Liu et al., 1997), apoptosis

predominantly affects oligodendrocytes after SCI (Casha and

Fehlings, 2001; Emery et al., 1998; Li et al., 1996; Liu et

al., 1997). Our data indicate that FAS inhibition may be

effective in decreasing this secondary injury mechanism and

improving neurological outcome.

The FAS death receptor pathway presents four sites in

particular which could serve as therapeutic targets for

pharmacotherapy development. These are FasL, the FAS

receptor, the initiator caspase 8 and the effector caspase 3.

Inhibition of the FasL and FAS receptor interaction has been

accomplished using neutralizing antibodies to these molecules

or by administering soluble FAS receptor (Demjen et al.,

2004; Schmidt et al., 2001; Silvestris et al., 2000). Inhibition

of the caspases has been accomplished through the use of

competitive and non-competitive inhibitors based on their

tetrapeptide cleavage specificity in animal models of brain

injury and ischemia (Clark et al., 2000; Gillardon et al., 1999;

Hara et al., 1997; Morita-Fujimura et al., 1999; Yakovlev et

al., 1997). Inhibition of signal transduction by manipulation

of molecules such as FADD and FLICE (caspase 8) inhibitory

protein (FLIP) has not yet been explored. Given what we

know from the FAS-deficient mouse, in vivo approaches to

inhibit the FAS pathway can be expected to have a profound

effect on the immune system possibly resulting in a

lymphoproliferative disorder and autoimmunity. This may

limit application of such strategies.

Other programmed cell death mechanisms are implicated

after SCI. These include the tumor necrosis factor death

receptor (TNFR), several caspases, several Bcl2 family

proteins, p53, cyclin D1, Cdk4 and the map kinase pathway

(Citron et al., 2000; Emery et al., 1998; Keane et al., 2001; Kim

et al., 2001; Lee et al., 2000; Li et al., 2000b; Nakahara et al.,

1999; Saito et al., 2000; Sakurai et al., 2000; Springer et al.,

1999). In addition, glutamate excitotoxicity, the inflammatory

response, free radicals and calpain activation contribute to post-

traumatic apoptosis in the spinal cord (Ghirnikar et al., 2001;

Ray et al., 2000a,b; Satake et al., 2000; Wada et al., 1999). Our

data suggest that these pathways also represent potential targets

in the treatment of SCI.

Acknowledgments

This work was funded by: The Canadian Institutes for

Health research, The Cervical Spine Research Society, The

Ontario Neurotrauma Foundation, The Christopher Reeve

Paralysis Foundation, Physicians Services Incorporated

Foundation (Ontario), Krembil Chair in Neural Repair and

Regeneration. The authors wish to thank Yang Liu and Amy

Lem for excellent technical assistance.

S. Casha et al. / Experimental Neurology 196 (2005) 390–400 399

References

Abe, Y., Yamamoto, T., Sugiyama, Y., Watanabe, T., Saito, N., Kayama, H.,

Kumagai, T., 1999. Apoptotic cells associated with Wallerian degeneration

after experimental spinal cord injury: a possible mechanism of oligoden-

droglial death. J. Neurotrauma 16, 945–952.

Basso, D.M., Beattie, M.S., Bresnahan, J.C., 1995. A sensitive and reliable

locomotor rating scale for open field testing in rats. J. Neurotrauma 12,

1–21.

Bjartmar, C., Yin, X., Trapp, B.D., 1999. Axonal pathology in myelin

disorders. J. Neurocytol. 28, 383–395.

Casha, S., Yu, W.R., Fehlings, M.G., 2001. Oligodendroglial apoptosis occurs

along degenerating axons and is associated with FAS and P75 expression

following spinal cord injury. Neuroscience 103, 203–218.

Citron, B.A., Arnold, P.M., Sebastian, C., Qin, F., Malladi, S., Ameenuddin, S.,

Landis, M.E., Festoff, B.W., 2000. Rapid upregulation of caspase-3 in rat

spinal cord after injury: mRNA, protein, and cellular localization correlates

with apoptotic cell death. Exp. Neurol. 166, 213–226.

Clark, R.S., Kochanek, P.M., Watkins, S.C., Chen, M., Dixon, C.E., Seidberg,

N.A., Melick, J., Loeffert, J.E., Nathaniel, P.D., Jin, K.L., Graham, S.H.,

2000. Caspase-3 mediated neuronal death after traumatic brain injury in

rats. J. Neurochem. 74, 740–753.

Crowe, M.J., Bresnahan, J.C., Shuman, S.L., Masters, J.N., Beattie, M.S., 1997.

Apoptosis and delayed degeneration after spinal cord injury in rats and

monkeys. Nat. Med. 3, 73–76.

Demjen, D., Klussmann, S., Kleber, S., Zuliani, C., Stieltjes, B., Metzger, C.,

Hirt, U.A., Walczak, H., Falk, W., Essig, M., Edler, L., Krammer, P.H.,

Martin-Villalba, A., 2004. Neutralization of CD95 ligand promotes

regeneration and functional recovery after spinal cord injury. Nat. Med. 7, 7.

Emery, E., Aldana, P., Bunge, M.B., Puckett, W., Srinivasan, A., Keane, R.W.,

Bethea, J., Levi, A.D., 1998. Apoptosis after traumatic human spinal cord

injury. J. Neurosurg. 89, 911–920.

Ghirnikar, R.S., Lee, Y.L., Eng, L.F., 2001. Chemokine antagonist infusion

promotes axonal sparing after spinal cord contusion injury in rat. Neurosci.

Res. 64, 582–589 (Jun 15).

Gillardon, F., Kiprianova, I., Sandkuhler, J., Hossmann, K.A., Spranger, M.,

1999. Inhibition of caspases prevents cell death of hippocampal CA1

neurons, but not impairment of hippocampal long-term potentiation

following global ischemia. Neuroscience 93, 1219–1222.

Hara, H., Friedlander, R.M., Gagliardini, V., Ayata, C., Fink, K., Huang, Z.,

Shimizu-Sasamata, M., Yuan, J., Moskowitz, M.A., 1997. Inhibition of

interleukin 1beta converting enzyme family proteases reduces ischemic

and excitotoxic neuronal damage. Proc. Natl. Acad. Sci. U. S. A. 94,

2007–2012.

Joshi, M., Fehlings, M.G., 2002a. Development and characterization of a novel,

graded model of clip compressive spinal cord injury in the mouse: Part 1.

Clip design, behavioral outcomes, and histopathology. J. Neurotrauma 19,

175–190.

Joshi, M., Fehlings, M.G., 2002b. Development and characterization of a novel,

graded model of clip compressive spinal cord injury in the mouse: Part 2.

Quantitative neuroanatomical assessment and analysis of the relation-

ships between axonal tracts, residual tissue, and locomotor recovery.

J. Neurotrauma 19, 191–203.

Katoh, K., Ikata, T., Katoh, S., Hamada, Y., Nakauchi, K., Sano, T., Niwa, M.,

1996. Induction and its spread of apoptosis in rat spinal cord after

mechanical trauma. Neurosci. Lett. 216, 9–12.

Keane, R.W., Kraydieh, S., Lotocki, G., Bethea, J.R., Krajewski, S., Reed, J.C.,

Dietrich, W.D., 2001. Apoptotic and anti-apoptotic mechanisms following

spinal cord injury. J. Neuropathol. Exp. Neurol. 60, 422–429.

Kim, G.M., Xu, J., Xu, J., Song, S.K., Yan, P., Ku, G., Xu, X.M., Hsu, C.Y.,

2001. Tumor necrosis factor receptor deletion reduces nuclear factor-

kappaB activation, cellular inhibitor of apoptosis protein 2 expression, and

functional recovery after traumatic spinal cord injury. J. Neurosci. 21,

6617–6625.

Krassioukov, A.V., Ackery, A., Schwartz, G., Adamchik, Y., Liu, Y., Fehlings,

M.G., 2002. An in vitro model of neurotrauma in organotypic spinal cord

cultures from adult mice. Brain Res. Brain Res. Protoc. 10, 60–68.

Lee, Y.B., Yune, T.Y., Baik, S.Y., Shin, Y.H., Du, S., Rhim, H., Lee, E.B., Kim,

Y.C., Shin, M.L., Markelonis, G.J., Oh, T.H., 2000. Role of tumor necrosis

factor-alpha in neuronal and glial apoptosis after spinal cord injury. Exp.

Neurol. 166, 190–195.

Li, G.L., Brodin, G., Farooque, M., Funa, K., Holtz, A., Wang, W.L., Olsson,

Y., 1996. Apoptosis and expression of Bcl-2 after compression trauma to rat

spinal cord. J. Neuropathol. Exp. Neurol. 55, 280–289.

Li, G.L., Farooque, M., Holtz, A., Olsson, Y., 1999. Apoptosis of

oligodendrocytes occurs for long distances away from the primary

injury after compression trauma to rat spinal cord. Acta Neuropathol.

(Berl) 98, 473–480.

Li, G.L., Farooque, M., Olsson, Y., 2000a. Changes of Fas and Fas ligand

immunoreactivity after compression trauma to rat spinal cord. Acta

Neuropathol. (Berl) 100, 75–81.

Li, X., Oudega, M., Dancausse, H.A., Levi, A.D., 2000b. The effect of

methylprednisolone on caspase-3 activation after rat spinal cord transection.

Restor. Neurol. Neurosci. 17, 203–209.

Liu, X.Z., Xu, X.M., Hu, R., Du, C., Zhang, S.X., McDonald, J.W., Dong,

H.X., Wu, Y.J., Fan, G.S., Jacquin, M.F., Hsu, C.Y., Choi, D.W., 1997.

Neuronal and glial apoptosis after traumatic spinal cord injury. J. Neurosci.

17, 5395–5406.

Lou, J., Lenke, L.G., Ludwig, F.J., O’Brien, M.F., 1998. Apoptosis as a

mechanism of neuronal cell death following acute experimental spinal cord

injury. Spinal Cord 36, 683–690.

Matsushita, K., Wu, Y., Qiu, J., Lang-Lazdunski, L., Hirt, L., Waeber, C.,

Hyman, B.T., Yuan, J., Moskowitz, M.A., 2000. Fas receptor and neuronal

cell death after spinal cord ischemia. J. Neurosci. 20, 6879–6887.

Morita-Fujimura, Y., Fujimura, M., Kawase, M., Murakami, K., Kim, G.W.,

Chan, P.H., 1999. Inhibition of interleukin-1beta converting enzyme family

proteases (caspases) reduces cold injury-induced brain trauma and DNA

fragmentation in mice. J. Cereb. Blood Flow Metab. 19, 634–642.

Nakahara, S., Yone, K., Sakou, T., Wada, S., Nagamine, T., Niiyama, T., Ichijo,

H., 1999a. Induction of apoptosis signal regulating kinase 1 (ASK1) after

spinal cord injury in rats: possible involvement of ASK1-JNK and -p38

pathways in neuronal apoptosis. J. Neuropathol. Exp. Neurol. 58, 442–450.

Nakatsuru, S., Terada, M., Nishihara, M., Kamogawa, J., Miyazaki, T., Qu,

W.M., Morimoto, K., Yazawa, C., Ogasawara, H., Abe, Y., Fukui, K.,

Ichien, G., Ito, M.R., Mori, S., Nakamura, Y., Nose, M., 1999b. Genetic

dissection of the complex pathological manifestations of collagen disease in

MRL/lpr mice. Pathol. Int. 49, 974–982.

Ray, S.K., Matzelle, D.C., Wilford, G.G., Hogan, E.L., Banik, N.L., 2000a. E-

64-d prevents both calpain upregulation and apoptosis in the lesion and

penumbra following spinal cord injury in rats. Brain Res. 867, 80–89.

Ray, S.K., Matzelle, D.D., Wilford, G.G., Hogan, E.L., Banik, N.L., 2000b.

Increased calpain expression is associated with apoptosis in rat spinal cord

injury: calpain inhibitor provides neuroprotection. Neurochem. Res. 25,

1191–1198.

Saito, N., Yamamoto, T., Watanabe, T., Abe, Y., Kumagai, T., 2000.

Implications of p53 protein expression in experimental spinal cord injury.

J. Neurotrauma 17, 173–182.

Sakurai, M., Hayashi, T., Abe, K., Sadahiro, M., Tabayashi, K., 1998. Delayed

selective motor neuron death and fas antigen induction after spinal cord

ischemia in rabbits. Brain Res. 797, 23–28.

Sakurai, M., Hayashi, T., Abe, K., Itoyama, Y., Tabayashi, K., Rosenblum,

W.I., 2000. Cyclin D1 and Cdk4 protein induction in motor neurons after

transient spinal cord ischemia in rabbits. Stroke 31, 200–207 (Jan.)

Satake, K., Matsuyama, Y., Kamiya, M., Kawakami, H., Iwata, H., Adachi,

K., Kiuchi, K., 2000. Nitric oxide via macrophage iNOS induces

apoptosis following traumatic spinal cord injury. Brain Res. Mol. Brain

Res. 85, 114–122.

Schmidt, M., Lugering, N., Lugering, A., Pauels, H.G., Schulze-Osthoff,

K., Domschke, W., Kucharzik, T., 2001. Role of the CD95/CD95 ligand

system in glucocorticoid-induced monocyte apoptosis. J. Immunol. 166,

1344–1351.

Schumacher, P.A., Eubanks, J.H., Fehlings, M.G., 1999. Increased calpain

I-mediated proteolysis, and preferential loss of dephosphorylated NF200,

following traumatic spinal cord injury. Neuroscience 91, 733–744.

Schumacher, P.A., Siman, R.G., Fehlings, M.G., 2000. Pretreatment with

calpain inhibitor CEP-4143 inhibits calpain I activation and cytoskeletal

S. Casha et al. / Experimental Neurology 196 (2005) 390–400400

degradation, improves neurological function, and enhances axonal survival

after traumatic spinal cord injury. J. Neurochem. 74, 1646–1655.

Shuman, S.L., Bresnahan, J.C., Beattie, M.S., 1997. Apoptosis of microglia and

oligodendrocytes after spinal cord contusion in rats. J. Neurosci. Res. 50,

798–808.

Siegel, R.M., Chan, F.K., Chun, H.J., Lenardo, M.J., 2000. The multifaceted

role of Fas signaling in immune cell homeostasis and autoimmunity. Nat.

Immunol. 1, 469–474.

Silvestris, F., Cafforio, P., Tucci, M., Del Prete, A., Dammacco, F., 2000.

VEINCTR-N, an immunogenic epitope of Fas (CD95/Apo-I), and soluble

Fas enhance T-cell apoptosis in vitro. II. Functional analysis and possible

implications in HIV-1 disease. Mol. Med. 6, 509–526.

Springer, J.E., Azbill, R.D., Knapp, P.E., 1999. Activation of the caspase-3

apoptotic cascade in traumatic spinal cord injury. Nat. Med. 5, 943–946.

Ugolini, G., Raoul, C., Ferri, A., Haenggeli, C., Yamamoto, Y., Salaun, D.,

Henderson, C.E., Kato, A.C., Pettmann, B., Hueber, A.O., 2003. Fas/tumor

necrosis factor receptor death signaling is required for axotomy-induced

death of motoneurons in vivo. J. Neurosci. 23, 8526–8531.

Vidal, S., Kono, D.H., Theofilopoulos, A.N., 1998. Loci predisposing to

autoimmunity in MRL-Fas lpr and C57BL/6-Faslpr mice. J. Clin. Invest.

101, 696–702.

Wada, S., Yone, K., Ishidou, Y., Nagamine, T., Nakahara, S., Niiyama, T.,

Sakou, T., 1999. Apoptosis following spinal cord injury in rats and

preventative effect of N-methyl-d-aspartate receptor antagonist. J. Neuro-

surg. 91, 98–104.

Warden, P., Bamber, N.I., Li, H., Esposito, A., Ahmad, K.A., Hsu, C.Y., Xu,

X.M., 2001. Delayed glial cell death following wallerian degeneration in

white matter tracts after spinal cord dorsal column cordotomy in adult rats.

Exp. Neurol. 168, 213–224.

Yakovlev, A.G., Knoblach, S.M., Fan, L., Fox, G.B., Goodnight, R., Faden,

A.I., 1997. Activation of CPP32-like caspases contributes to neuronal

apoptosis and neurological dysfunction after traumatic brain injury.

J. Neurosci. 17, 7415–7424.

Yin, X., Crawford, T.O., Griffin, J.W., Tu, P., Lee, V.M., Li, C., Roder, J.,

Trapp, B.D., 1998. Myelin-associated glycoprotein is a myelin signal that

modulates the caliber of myelinated axons. J. Neurosci. 18, 1953–1962.

Yong, C., Arnold, P.M., Zoubine, M.N., Citron, B.A., Watanabe, I., Berman,

N.E., Festoff, B.W., 1998. Apoptosis in cellular compartments of rat spinal

cord after severe contusion injury. J. Neurotrauma 15, 459–472.

Yoshino, O., Matsuno, H., Nakamura, H., Yudoh, K., Abe, Y., Sawai, T., Uzuki,

M., Yonehara, S., Kimura, T., 2004. The role of Fas-mediated apoptosis

after traumatic spinal cord injury. Spine 29, 1394–1404.

Zurita, M., Vaquero, J., Zurita, I., 2001. Presence and significance of CD-95

(Fas/APO1) expression after spinal cord injury. J. Neurosurg. 94, 257–264.